The Potential of the Dynamic Transmission Electron Microscope (DTEM

Document Sample
The Potential of the Dynamic Transmission Electron Microscope (DTEM Powered By Docstoc
					             The Potential of the Dynamic
          Transmission Electron Microscope
            (DTEM) for In Situ Microscopy

                                        CMS
                                               chemistry &
                                               materials
                                               science




                             Presented by Bryan W. Reed
                                Monday, August 7, 2006




Michael Armstrong, Ken Boyden, Nigel D. Browning, Geoffrey
  H. Campbell, Jeffrey D. Colvin, Bill DeHope, Alan M. Frank,
David J. Gibson, Fred Hartemann, Judy S. Kim, Wayne E. King,
Thomas B. LaGrange, Ben J. Pyke, Bryan W. Reed, Richard M.
        Shuttlesworth, Brent C. Stuart, Ben R. Torralva
                          University of California
                   Lawrence Livermore National Laboratory
   Work performed under the auspices of the U.S. Department of Energy by University of California,
              Lawrence Livermore National Laboratory under Contract W-7405-Eng-48
    This work is funded in part by the Office of Basic Sciences of the U. S. Department of Energy
                                                     CMS
                                                      chemistry &
                                                      materials

Plan of Presentation                                  science




• Part I: Motivation, Context, Overview

• Part II:    Current Results and Applications


• Part III:   Performance and Future Directions




                                UCRL-PRES-222787.2
High temporal resolution affords access to
                                                             CMS
                                                              chemistry &
fundamental processes in Biology, Chemistry,                  materials
                                                              science

and Materials Science




      Many studies would benefit from high time resolution

                                        UCRL-PRES-222787.3
                                                    CMS
                                                        chemistry &
UED and DTEM have roots extending back to               materials
                                                        science

the mid 1970s and beyond

                                                      UED
                                                    Ultrafast
                                                    Electron
                                                   Diffraction

                                                     DTEM
                                                    Dynamic
                                                  Transmission
                                                    Electron
                                                   Microscopy




                             UCRL-PRES-222787.4
                                                                      CMS
                                                                       chemistry &
     Dynamic TEM (DTEM) fits into a broad context                      materials
                                                                       science

     of ultrafast and in situ science

•   Like in-situ TEM, but faster
     – Nanoseconds instead of (typically) video rate
     – Very few frames per movie

•   Like ultrafast x-ray probes, but with electrons
     – Much higher interaction cross sections
     – Much less radiation damage per elastic event
     – Much higher typical imaging resolution
     – Less coherence, difficult to do spectroscopy

•   Like ultrafast electron diffraction (UED), but with imaging and
    small-area selectivity
     – Very flexible; many operating modes; single-shot operation
     – Sacrifice of simplicity and time resolution

                                               UCRL-PRES-222787.5
                                                                   CMS
                                                                        chemistry &
DTEM combines short pulses with traditional                             materials
                                                                        science
electron optics

 •   JEOL 2000FX
 •   Standard thermionic filament
     replaced by photocathode                    Cathode      Photocathode
                                                  Laser
 •   Access provided for                          Drive
     photocathode laser                           Laser        Sample

 •   Access provided for drive laser     Variable
                                          Delay
 •   “Every electron” camera
 •   Future: “Movie-mode” fast
     deflector
 •   Different operating modes:           Camera
      – ns/nm single shot
      – µs/Angstrom single shot
      – ps/Angstrom stroboscopic       LLNL Dynamic Transmission Electron
                                                 Microscope


                                         UCRL-PRES-222787.6
                                                                  CMS
                                                                       chemistry &
The sample is driven with a synchronized laser                         materials
                                                                       science

pulse

• Direct illumination of sample possible at ≈35˚
• Focused spot ≈ 30-70 µm




           Lens outside
           of column                                             Electron probe
                                                                 arrives 0-5000 ns
                                                                 later (pump-probe)




                                            UCRL-PRES-222787.7
                                                                   CMS
                                                                        chemistry &
                                                                        materials
                                                                        science

    How do you make a movie?
Three methods:                           Streak Al melting
                                         Bostanjoglo and Nink,
•   Streak                               JAP 79, 8725 (1996)
     – This gives you a 1-D image at every
        point in time
•   A fast deflector that can raster multiple
    images on one large camera frame
     – This is what TU-Berlin did, with up to 3                    IL
        frames per movie, switching time ~20 ns
     – An engineering challenge in high-speed,                     PL
        high-voltage switching
•   A fast-framing camera
     – One CCD chip that can locally store
        multiple images
     – Switching among 4-300 frames at 1-100
        MHz is now possible (e.g. commercial
        units from Princeton Scientific and DRS
        Technologies)
                                              UCRL-PRES-222787.8
                                                     CMS
                                                      chemistry &
                                                      materials

Plan of Presentation                                  science




• Part I:    Motivation, Context, Overview


• Part II:    Current Results and Applications


• Part III:   Performance and Future Directions




                                UCRL-PRES-222787.9
                                                                    CMS
                                                                         chemistry &
Experiments demonstrate 20 nm/30 ns single-                              materials
                                                                         science

shot resolution




                                                    Thomas LaGrange et al.


   Lengthening the UV laser pulse duration from 3 ns to 30 ns has
   increased signal levels and image quality.
   Better than 20 nm spatial resolution using 30 ns electron pulses.




                                         UCRL-PRES-222787.10
                                                            CMS
                                                                    chemistry &
                                                                    materials
                                                                    science

Assessing the resolution of single shot images



           Line Scan and
                                      30 nm
           Average                    thick layers


                                                     20 nm
                                                     thick layers




 20 nm line/20 nm space
 20 nm line/20 nm space
modulations exceed noise
modulations exceed noise



                               UCRL-PRES-222787.11
                                                                CMS
                                                                   chemistry &
Effect of Electron Pulse Duration on Diffraction                   materials
                                                                   science

Pattern Resolution




Continuous-Wave               30 ns                         1.5 ns
      (CW )              electron pulse                 electron pulse

            We optimize time resolution against required
            We optimize time resolution against required
                 beam quality for each experiment
                 beam quality for each experiment

                                      UCRL-PRES-222787.12
                                                         CMS
                                                          chemistry &
                                                          materials
                                                          science

Example applications

• α_β transition in pure Ti thin films

• Reactive multi-layer foils

• Coordinated molecular/crystallographic phase changes

• Biological materials


                                   UCRL-PRES-222787.13
                                                                           CMS
                                                                                 chemistry &
The α to β transition in Ti occurs by Martensitic                                materials
                                                                                 science

type transformation.
•   Occurs upon heating to above 1155K
    (882°C)
•   A simple shear of the HCP structure
    generates the BCC                     HCP
     – [111] (112) or [111] (101)
•   Habit plane is (433) β

•   Lattice correspondence
     – {0001}α ll {110} β
     – 〈1120〉 α ll 〈111〉β
•   6 BCC crystallographic variants are
    possible from 1 HCP crystal           BCC


                                                     Transformation path from Zhang,
                                                      et al, APL, Vol. 84, no. 24 (2004)




                                          UCRL-PRES-222787.14
                                                                CMS
                                                                     chemistry &
First dynamic experiment on the DTEM successfully                    materials
                                                                     science

demonstrated in situ observation of ultrafast phenomena

                                             Melting transition




                                    hcp Ti, α-phase       Melt structure

                                      Phase transformation (α to β)




                                    hcp Ti, α-phase       bcc Ti, β-phase

                                    UCRL-PRES-222787.15
Phase fractions are calculated by rotationally averaging
                                                                 CMS
                                                                     chemistry &
                                                                     materials
difference patterns and by determining the integrated                science

intensities from peak fitting routines (Rietveld Refinement).




                                                   250ns, 29% BCC phase
                                                   500ns, 50% BCC phase




              results contained in article submitted to APL
              results contained in article submitted to APL

                                         UCRL-PRES-222787.16
                                                             CMS
                                                                   chemistry &
The α_β transition in Ti observed through a                        materials
                                                                   science

series of diffraction patterns


                                             Movie built up from series of
                                             pump-probe experiments in a
                                             single-crystal region that could
                                             be driven repeatably.


                                             Transforms hcp to bcc and back.


                                             Parent-daughter orientation
                                             relationship apparent.




                                UCRL-PRES-222787.17
                                                        CMS
                                                             chemistry &
Grain growth occurs in the films due to the laser            materials
                                                             science

irradiation and annealing processes




                           ?
                                                      LATH




                                     LATH




                                                      LATH



                                     After laser treatment
  Before laser treatment
                                     100-500 nm grain size
  60 nm equi-axed grains
                                     Mixed microstructure
                                     lath + small ‘pancake’
                                UCRL-PRES-222787.18
                                                                                         CMS
                                                                                               chemistry &
                                                                                               materials
                                                                                               science

    Application Area: Reactive Multilayer Foils
                                                                Application
•   Repeated metal bi-layers
                                                                • Fuse dissimilar materials, exposing
•   Exothermically form intermetallics                            only superficial material to heat

•   Self-propagating reaction forms new                         Interest
    phases                                                      • Metastable states cannot be
              Propagation Direction                                observed conventionally
                                                                • Microscopic details are unknown
        Reacted Foil                   Unreacted Foil
                                                                                               Ni / Cu
Spark
                                                                                                 Al
                       Reaction Zone


                                        Atomic
         Atomic                         Diffusion
         Mixing
                                                                                                10 nm
                                                        Ni
                                                        Al
                  Thermal Diffusion                                      Unreacted bi-layers
                                                        Intermetallic
                                                            UCRL-PRES-222787.19
                                                                     CMS
                                                                         chemistry &
                                                                         materials
                                                                         science

DTEM In situ Study

     laser




 Intermetallics form radially
 outward from initiating laser
 spot until entire foil has been
 reacted

                                                               Separate phases
                                                               of Al and Ti

                                   Intermetallic Al2Ti

                                         UCRL-PRES-222787.20
                                                                                                                  CMS
                                                                                                                      chemistry &
First DTEM results: New crystal phases don’t                                                                          materials
                                                                                                                      science
appear until long after the reaction front passes

                                                                                               Upon arrival of the front, the
                       Al{200} and Ni {111}                                                    peaks shift left and weaken at
            Al {111}


                                              Ni {200}                                         high k values, but don't show a
                                                                                               transformation to a new phase
                                                         Al {220}

                                                                    Ni{220}


                                                                              Ni{311}
                                                                                               until much later.

                                                                                               Real-space image shows
                                                                                               significant morphology change
                                                                                               happens almost right away.




   Al/Ni0.93V0.07 foil 200 µm from initiation




                                                                                        UCRL-PRES-222787.21
                                                                                       CMS
                                                                                        chemistry &
Application Area: Molecular phase                                                       materials
                                                                                        science

transformation in HMX

  “Chair”                                                                          “Boat”
conformation                                                                    conformation
                                       170 Celcius


                                        7% volume
                                         change
Monoclinic                                                                     Hexagonal



                        β−HMX                                  δ−HMX
   •   Significant molecular conformation change upon phase transformation

   •   Nucleation and growth controlled by physics on multiple length scales

   •   Current data indicate a large activation energy for nucleation – is molecular
       coordination a barrier to nucleation?

   •   The atomic-scale details of molecular phase transformations are virtually unknown –
       intermediate phases?, “virtual melting”?
                                                        UCRL-PRES-222787.22
                                                                         CMS
                                                                             chemistry &
DTEM has obtained adequate-quality single-shot diffraction                   materials
                                                                             science
patterns on HMX with negligible radiation damage


                                       •   Spots can be resolved and pattern
                                           indexed despite large unit cell and
                                           electron beam convergence
                                       •   Signal sufficient despite low atomic
                                           number and high radiation sensitivity
                                       •   Same sample region can be probed many
                                           times with minimal damage
                                             – This was difficult to achieve in
                                               standard TEM


 [100] β-HMX single-shot diffraction




                                                   UCRL-PRES-222787.23
                                                                                   CMS
                                                                                        chemistry &
  We have recently demonstrated single-shot                                             materials
                                                                                        science

  diffraction from crystalline paraffin

                                  CW                         Pulsed




                                                                            Single shot diffraction
                                                                            pattern of a monolayer of
                                                                            orthorhombic C44H90
                                                                            paraffin obtained on the
                                                                            DTEM




Michael Armstrong and Thomas LaGrange, LLNL

    •   Radiation damage limits high resolution imaging of biosamples
    •   What is the time scale for damage? Is a nanosecond pulse short enough?
    •   Crystalline paraffin is an ideal sample to test this hypothesis



                                                      UCRL-PRES-222787.24
                                                      CMS
                                                       chemistry &
                                                       materials

Plan of Presentation                                   science




• Part I: Motivation, Context, Overview

• Part II:    Current Results and Applications


• Part III:   Performance and Future Directions




                                UCRL-PRES-222787.25
                                                                   CMS
                                                                         chemistry &
                                                                         materials
                                                                         science

What's important to building a good DTEM?
•   Optimal gun design
     – Laser parameters
     – Materials
     – Accelerating and focusing fields
     – Making the most of the brightness you have
     – Best design may be very different from conventional designs
•   Avoid electron-electron effects
     – Minimize impact of harmful crossovers
     – Operate at high voltage
•   Be able to operate at relatively high emittance (low beam quality)
     – Choice of experiments
     – Choice of apertures, operating conditions, etc.
     – Low-aberration lenses
•   Have a flexible, high-performance laser system
•   Have a good camera


                                             UCRL-PRES-222787.26
    How many electrons do you need in a pulse?
                                                                         CMS
                                                                            chemistry &
                                                                            materials
                                                                            science




•    Imaging mode: 106-108 electrons
     detected at the camera
•    Diffraction mode: 105-107
•    Pump-probe accumulation or        Original          104 electrons   105 electrons
     "stroboscopic" mode
      – This was the first "DTEM"
         with Δt = 200 ps
         (Bostanjoglo, 1976)
      – Limited to viewing things      106 electrons
                                       106 electrons     107 electrons
                                                         107 electrons   1078 electrons
                                                                         10 electrons
         you can drive reversibly/
         repeatably many times                Simulation, 512 x 512 image


                  Rule #1 of DTEM: Electrons are precious.
                   This determines your choice of source,
                  apertures, operating modes, and cameras.

                                             UCRL-PRES-222787.27
                                                                          CMS
                                                                           chemistry &
                                                                           materials
                                                                           science

     Brightness limits time resolution
                                                              Ne
Brightness = current per area per solid angle =
                                                       (πr 2 )(πα 2 )Δt
Let N = 108 electrons, r = 1 µm, α = 10 mrad, Δt = 1 ns.
Brightness = 1.6x109 A cm-2 steradian-1, with 16 mA at the camera.

                   Photoemission is the only option.
              Field emission will not achieve this current.
              Thermionic will not achieve this brightness.
        Neither can easily be turned on and off in a nanosecond.
          Time resolution in DTEM is controlled by laser pulse length
         and determined by brightness and experimental requirements.

         Time resolution comes at the expense of coherence.
     Large source sizes, convergence angles, and energy spreads.
     (Higher accelerating voltages help by increasing brightness.)

                                                 UCRL-PRES-222787.28
                                                                     CMS
                                                                      chemistry &
    Electron-electron effects can limit the current                   materials
                                                                      science

    and coherence
                                                          Filament
•   Electric field force due to
    cathode-anode potential

•




                                                                          lt
    Additional field due to




                                                                        ne
    Wehnelt bias




                                                                      eh
                                                                     W
•   Normal crossover formation
     – Boersch effect reduces
       temporal coherence
                                        Good gun design for CW mode:
•   Space charge blowup                 Small emission area
•   Additional field due to recently-   Tight crossover
    emitted electrons                   Low electric fields for high stability
     – Child's law reduces current
     – Lateral spreading reduces
       spatial coherence                Good gun design for pulsed mode:
     – Gun crossover is weakened        Almost exactly the opposite
       or eliminated

                                            UCRL-PRES-222787.29
                                                                                                                     CMS
                                                                                                                         chemistry &
Electron yield shows a saturation effect as the                                                                          materials
                                                                                                                         science

cathode laser energy increases

 Linear part shows QE                                           Yield vs. laser energy                        Saturation
 ~1% of expected                                                                                            Space charge?
                                              1.6
                                                                                                            Voltage droop?
                                              1.4                                                           Charge limit?
                 Yield (# of electrons/106)


                                              1.2
                                               1

                                              0.8

                                              0.6

                                              0.4
                                              0.2

                                               0
                                                    0   2   4        6       8       10     12     14
                                                                  Input Laser Energy (µJ)



Changing parameters rescales the axes but hardly changes the shape of the curve.

         - What limits the yield in the linear part?
         - What limits the yield in the saturation regime?


                                                                                      UCRL-PRES-222787.30
  Electrostatic simulations suggest that
                                                                                        CMS
                                                                                            chemistry &
  we must trade off divergence against                                                      materials
                                                                                            science

  cathode electric field
Cathode flush with wehnelt               Cathode 0.3 mm behind wehnelt

10 MV/m at cathode center                5 MV/m at cathode center

>90% of electrons blocked                Much less divergence
                              Wehnelt
                              Cathode                                            Must re-optimize
                                                                                 µA electron gun
                                                                                  to work well at
                                                                                     milliamps
                              Electron
                               Tracks                                   •    Cathode/wehnelt geometry
                                                                        •    Wehnelt bias
                                                                        •    Laser spot size/shape




                              1 mm
                             Aperture
                                                           UCRL-PRES-222787.31
  Electron-electron interactions affect
                                                             CMS
                                                                 chemistry &
                                                                 materials
                                                                 science
  performance at high currents

108 electrons/ns @ 200 keV = one electron every 2 nm down the column

        Homogeneous Effects                Inhomogeneous Effects
"Space charge"                        Includes Boersch effect
Continuum model                       Discrete, stochastic model
ρ, v, and E are continuous            Localized, random individual
functions of space and time           charges and statistical collisions




            Electron-electron repulsion is our worst enemy

                                       UCRL-PRES-222787.32
   Electron-electron interactions come
                                                                            CMS
                                                                                chemistry &
                                                                                materials
                                                                                science
   in multiple flavors
   Longitudinal Space Charge                          Lateral Space Charge
   Cathode    Child's Law:
                                                                Defocusing:
              • Field from recent
      ρ                                                         • To first order, it's just
           qE    electrons retards
                                                                   a diverging lens
                 emission
                                                                • Higher orders create
              • Fundamentally limits
Anode   Anode                                                      spherical aberration
                 current density:
                            j ∝ V 3/ 2 / d 2
 Longitudinal Inhomogeneous                       Lateral Inhomogeneous
                           Boersch Effect:                   Trajectory Displacement:
                                                             • Random electron-
             Propagation




                           • Pulse viewed
                              in rest frame                      electron scatter
                           • Most velocity                   • High-resolution
                              is lateral                         information stored in
                                                                 lateral velocities is
                           • Equilibration
                                                                 gradually destroyed
                              increases ΔE


                                               UCRL-PRES-222787.33
                                                                  CMS
                                                                       chemistry &
                                                                       materials
                                                                       science
More than one electron is in the sample at a time!

        16 mA @ 200 keV           = 108 electrons per ns
                                  = 1 electron per 10 as
                                  = one electron per 2 nm
   All of these electrons are creating plasmons and repelling each other.
      Can we still use conventional TEM image contrast theory?

 108 electrons in a 100 nm spot    = 1 electron per Å2
                                   = 1 eV deposited per H2O molecule

 In 1 ns, a low conductivity sample might only dissipate 10% of this energy
 to the surroundings.


   What are the sample damage mechanisms for DTEM intensities?

                                            UCRL-PRES-222787.34
                                                                          CMS
                                                                              chemistry &
    Conventional resolution limits are not                                    materials
                                                                              science

    necessarily the deciding factor for DTEM
•   Observed resolution limits
    ~150 nm (TU-Berlin), ~15 nm
    (LLNL)
•   Measured energy spread ~10
    eV (TU-Berlin)
•   Assumed convergence angles
    are ~10 milliradians
•   Calculations indicate
    conventional resolution limit
    ~0.5-3 nm (amplitude contrast)
•   Atomic resolution is unlikely

                            Results calculated using Java EMS,
                            http://cimewww.epfl.ch/people/stadelmann/jemswebsite/jems.html


                                                 UCRL-PRES-222787.35
                                                                CMS
                                                                 chemistry &
Post-sample trajectory displacement is probably                  materials
                                                                 science

the main resolution limitation
         Sample
                                            Near a crossover:
              Lens
             Crossover     •   Intensity is very high
               Plane       •   Sample position information is encoded
                               primarily as lateral momentum
                           •   Anything that deranges lateral velocity
              Image            distribution blurs the final image
              Plane

                         Two such effects:
                  Lateral trajectory displacement
                   (~10-50 nm resolution limit)

          Space-charge induced spherical aberration
              (probably a much smaller effect)

                                        UCRL-PRES-222787.36
                                                                         CMS
                                                                               chemistry &
                                                                               materials
Various methods of estimation agree on a 10-50                                 science

nm trajectory displacement resolution limit
Trajectory displacement has been heavily studied for projection lithography.



                             IL3 PL
                       IL2
                 IL1
     OL




Analytical formulas were applied to optical   Numerical multi-particle simulation of
model of post-sample lens system.             electron point spread function after
Result: 10-50 nm resolution limit,            sample interaction
depending on current and lens settings        Result: 10-50 nm resolution limit,
                                              depending on current and spot size


                                                UCRL-PRES-222787.37
                                                                                       CMS
                                                                                           chemistry &
   Simulation has suggested the feasibility of                                             materials
                                                                                           science

   ultrafast (< 10 ps) electron imaging

       Point Spread Function at 200 keV, 1 ns         Point Spread Function at 5 MeV, 10 ps




     -100 nm   -50 nm    0     50 nm 100 nm        -100 nm   -50 nm      0      50 nm 100 nm


Both simulations have 1.6 x 105 electrons, showing progressive blurring to a 5 mm propagation distance.

        For picosecond resolution imaging, relativistic energies
         may be the only way to obtain high spatial resolution.

                                                          UCRL-PRES-222787.38
                                                                     CMS
                                                                           chemistry &
DTEM is crying out for engineering that                                    materials
                                                                           science

combines TEM and accelerator technology
                        DTEM requires      Conventional          Accelerators
                                              TEM
Current                      mA               nA to µA             A to kA

Voltage                 100 to 1000 keV     few 100 keV           Multi-MeV

Lateral emittance         nm-radian          nm-radian            µm-radian

Energy spread               10 eV            0.1 to 3 eV             keV

Time resolution            ps to µs              ms                   ps

Focal lengths and         mm to cm               mm                cm to m
principal aberrations
Aberration correction   Would be nice      Under rapid          Not in the same
                                           development               sense
Pulse compression       Would be great     Nonexistent             Routine




                                          UCRL-PRES-222787.39
                                                                       CMS
                                                                           chemistry &
                                                                           materials
   A future DTEM might look more like a particle                           science

   accelerator
                                 Integrated optical
                                   pump/probe at
                 Pulse                sample
                                                                 Fast framing single-
               compressor
                                                                  electron-sensitive
                                                                   camera system




RF photogun optimized
 for 0.01-1 A, 1-5 MV
                                                       Rapid
                                                deflection/streaking
             Complex multipole                        system
                  lenses
            (Superconducting?)

                    Heavy shielding.
                   Experimenter is not
                      in the room!         UCRL-PRES-222787.40
                                                                  CMS
                                                                     chemistry &
                                                                     materials
                                                                     science

Conclusions
•   Time-resolved electron microscopy is being rapidly developed
     – Roots go back to 1960's, but recently has picked up pace
     – Complements other techniques including pulsed x-ray and optical,
       ultrafast electron diffraction, and video-rate in-situ TEM
•   Need to trade off something to gain the time resolution
     – 20 nm/30 ns imaging resolution so far
     – Expect sub-nanometer (atomic?) resolution somewhere in µs regime
•   Applications in studying material behavior on µs and ns scales
     – Phase transformations, Dislocation dynamics, Surface physics, Organic
       materials
•   Getting it to work requires bringing together many disciplines
     – TEM, Lasers, Accelerator Physics, Materials Science, Modeling and
       Simulation
•   In the future, we may see picosecond-scale few-MeV instruments that look
    very little like a conventional TEM


                                            UCRL-PRES-222787.41
                                                                   CMS
                                                                    chemistry &
DTEM was developed at TU-Berlin beginning in                        materials
                                                                    science

the late 1970's
PI Oleg Bostanjoglo
•   1976 stroboscopic mode @ 2 ps
•   1987 single-pulse images @ 20 ns
•   By 2003: 7 ns, 3-frame movies,
    diffraction patterns, streak imaging,
    nanosecond PEEM, . . .
•   Based on Siemens Elmiskop IA




                           H. Dömer and O.
                           Bostanjoglo, Rev. Sci.
                           Inst. 74, 4369 (2003)

                                             UCRL-PRES-222787.42
                                                                                           CMS
                                                                                             chemistry &
Dynamic TEM study of ablation of Ni film by                                                  materials
                                                                                             science

ultrashort laser pulse
      2.5 ns                                 5 ns                                    10 ns




      20 ns                                 40 ns                                      +∞
           H. Domer and O. Bostanjoglo, Journal of Applied Physics 91, 5462-5467 (2002).


                                                        UCRL-PRES-222787.43
                                                           CMS
                                                            chemistry &
The Zewail group at Caltech is also constructing            materials
                                                            science

at least one DTEM

•   120 keV G12 Twin Tecnai
•   Gatan Ultrascan 1000 UHS
    2048 x 2048
•   Results reported were acquired
    with 1 electron per pulse




       PNAS 2005;102[20]:7069-73




                                     UCRL-PRES-222787.44
                                                         CMS
                                                             chemistry &
DTEM has demonstrated <50nm resolution in                    materials
                                                             science

single shot mode




                            •    Contrary to recent literature
                                 comments, we have observed
                                 resolution far better than 1 µm in
                                 the LLNL DTEM

                                UCRL-PRES-222787.45
                                                                   CMS
                                                                         chemistry &
                                                                         materials
                                                                         science

UED systems are relatively simple




                                                     C. Y. Ruan, F. Vigliotti, V. A.
                                                     Lobastov, S. Y. Chen, and
                                                     A. H. Zewail, Proceedings of
                                                     the National Academy of
                                                     Sciences of the United
                                                     States of America 101,
                                                     1123-1128 (2004).




                               UCRL-PRES-222787.46
                                                                                                                     CMS
                                                                                                                      chemistry &
Currently, UED and DTEM have distinctly                                                                               materials
                                                                                                                      science

different operating characteristics

                                                        UED                                     DTEM

   Time resolution                                 300-600 fs                                  2-10 ns
   Electrons per image
                                                    105 - 1010                                 106 -108
   or diffraction pattern
   Electrons per shot                           Few thousand                                  106 - 108
   Shots per image or
                                                     102 - 107                                        1
   diffraction pattern
   Single shot capable?                               Not yet                                     Yes

   Movies?                                            Not yet                                     Yes

   Image Resolution                                      N/A                                 10-20 nm

       J. Cao, Z. Hao, H. Park, C. Tao, D. Kau, and L. Blaszczyk, Applied Physics Letters 83, 1044-1046 (2003).
       B. J. Siwick, J. R. Dwyer, R. E. Jordan, and R. J. D. Miller, Science 302, 1382-1385 (2003).
       P. M. Weber, R. C. Dudek, S. Ryu, and R. M. Startt, in Femtochemistry and Femtobiology: Ultrafast Events in
       Molecular Science, edited by M. M. Martin and J. T. Hynes (ELSEVIER, Amsterdam, 2004), p. 19-24.

                                                                           UCRL-PRES-222787.47
                                                                      CMS
                                                                       chemistry &
Experiments agree with calculations:                                   materials
                                                                       science

Millions of electrons per pulse are needed.

•   Every electron is precious
    and must be collected.
     – Capture images using a
         ‘single electron’
         sensitive CCD camera
     –   Proper choice of
         apertures, source size,
         etc. to optimize, signal,
         contrast, and resolution    5x108                    1x108          5x107
•   Imaging mode
    5x106-108 electrons
    ON THE CCD
•   Diffraction mode
    105-107 electrons
                                     1x107                    5x106          1x106




                                             UCRL-PRES-222787.48
                                                                               CMS
                                                                                     chemistry &
Ultrafast microscopy accesses the same time and                                      materials
                                                                                     science
length scales as modern computer simulations

                                                    Unshocked bcc iron

                                                     Shock-compressed
                                                     bcc iron

                                                  Transformed hcp iron

                                                   Studied using molecular dynamics at LANL.
                                                   Four different shock pressures shown.

                                                   Kadau, K., Germann, T. C., Lomdahl, P. S. &
                                                   Holian, B. L. Microscopic View of Structural
                                                   Phase Transitions Induced by Shock Waves.
                                                   Science 296, 1681-1684 (2002).


      Ultrafast Materials Science implies experiment at the scale of simulation
              Additional benefits if ultrafast imaging can be achieved

                                               UCRL-PRES-222787.49
                                                                                               CMS
                                                                                                      chemistry &
Modern RF photoguns used in particle                                                                  materials
                                                                                                      science

accelerators have some very promising specs
        Specifications                          Advantages


• Gun operates at 5.5 MeV          • “no” space charge
• 108 electrons per bunch          • Thicker targets
• 400 fs rms bunch length          • Single shot experiments
• 1 micron normalized emittance    • Improved time resolution
• 0.1 micron geometric emittance

                                    J. Schmerge, D. Dowell, and J. Hastings, "Measured RF gun parameters
                                    at the SLAC Gun Test Facility," presented to First National Lab and
                                    University Alliance Workshop on Ultrafast Electron Microscopies,
                                    Pleasanton, CA, April 16-17, 2004.




                                                       UCRL-PRES-222787.50
                                                                                          CMS
                                                                                              chemistry &
Single shot pattern with       electrons with a ~2x107                                        materials
                                                                                              science

bunch length of 500fs acquired at SLAC
         160 nm Foil in-beam                                           Foil out-of-beam




                                             (111), (200)     Dark current image subtracted
                                           (222)            Total bunch charge: 2 x 107
                                                            electrons
                                                    (311)   Aluminum foil thickness: 160 nm
                                                            Drift tube length: 3.95 m
   •   Jerry Hastings, Peter Weber, et al. of the           Beam Energy: 5.5 MeV
       SLAC and Brown University Groups
                                                            Pulse duration: 500 fs
                                                              UCRL-PRES-222787.51
                                                                            CMS
                                                                                  chemistry &
At higher currents, lateral space charge                                          materials
                                                                                  science

expansion limits collection efficiency

         Fringe Field Dominated
  Initial Diameter                     Initial Diameter




                                  108 electrons/ns = 16 mA
    Initial Diameter
                                  1 mm aperture at ~32 cm

                                  In space charge limit, adding electrons at cathode
                                  has little effect on current through aperture.

                                  Most lateral velocity appears in first few cm (fringing
                                  fields, electrons moving slowly at high density)



                                                   UCRL-PRES-222787.52
                                                              CMS
                                                                 chemistry &
                                                                 materials
                                                                 science

Future Studies

•       Diffraction Mode            •   Image Mode
         – Study transformation          – Study interface propagation
           kinetics, TTT diagrams
                                            • Directly observe if
                                              transformation is
                                              martensitic or massive
                                              type
    Temperature




                                            • Visualize transformation
                                              dislocation/ interfacial
                                              defect dynamics and
                                              elucidate transformation
                                              mechanisms

                     Time, ns




                                        UCRL-PRES-222787.53